基于改进决策树算法的失衡数据集分类方法

The Classification Method of Imbalanced Data Set Based on Improved Decision Tree Algorithm

为了提高云资源空间数据的检索能力,需要对云资源分布空间的失衡数据进行优化分类处理,提出基于改进决策树算法的失衡数据集分类算法,构建失衡数据分布的不规则空间聚类模型,采用特征空间重组方法进行失衡数据的模糊特征重构和聚类处理,提取失衡数据的关联特征分布集和属性集,根据失衡数据的属性分布进行大数据挖掘和自适应特征提取,采用改进决策树算法对提取的失衡数据特征集进行不规则三角网重构和模糊聚类处理,实现失衡数据的优化分类。仿真结果表明,采用该方法进行失衡数据分类的自动分类性能较好,失误率较低,提高了失衡数据的分类检测和识别能力。

In order to improve the retrieval capability of cloud resource spatial data,the imbalanced data of cloud resource distribution space needs to be optimized and classified.An imbalanced data set classification algorithm based on an improved decision tree algorithm is proposed to construct an irregular spatial clustering model of imbalanced data distribution.Feature space reorganization method performs fuzzy feature reconstruction and clustering processing of unbalanced data,extracts the associated feature distribution set and attribute set of unbalanced data,performs big data mining and adaptive feature extraction based on the attribute distribution of unbalanced data,and uses an improved decision tree algorithm.The irregular triangle network reconstruction and fuzzy clustering processing are performed on the extracted imbalanced data feature set to achieve the optimal classification of the imbalanced data.Simulation results show that the automatic classification performance of unbalanced data classification using this method is better,the error rate is lower,and the classification detection and recognition capabilities of unbalanced data are improved.